Identification and comparison of the markers of mortality in intensive care patients with septic or non-septic acute kidney injury

Introduction: Acute kidney injury (AKI) is characterized by an abrupt and progressive decline in glomerular filtration rate, in addition to accumulation of nitrogenous waste products and failure of the kidneys to maintain fluid/electrolyte and acid/base balances. AKI is common among intensive care patients and is associated with increased morbidity and mortality. Sepsis, on the other hand, is a serious clinical condition resulting from systemic response to infection and it involves a certain level of organ dysfunction. Both sepsis and AKI alone can result in significant mortality, whereas coexistence of these two entities leads to an even higher rate of mortality among the patients and a significant rate of morbidity among the survivors. Objective: Objectives of the present study include identification of the risk factors and laboratory parameters that can potentially affect the rates of mortality and morbidity, and to investigate the NLR and ΔNLR values as basic markers of inflammation in patients who are hospitalized in intensive care unit and develop AKI during follow-up. Methods: A total number of 201 patients, who were diagnosed with AKI while receiving inpatient treatment at Internal Diseases Department Intensive Care Unit of Ege University Medical Faculty Hospital between January 2012 – June 2014 and whose data could be retrospectively retrieved, were included in this study. Depending on the cause of AKI, all patients included in the study cohort (n=201) were assigned to one of the following groups: Group A, including the patients with septic AKI (n=112), and Group B, including the patients with AKI due to non-septic etiologies (n=89). Findings: Mean age of the patients included in this study was 72.5 ± 14.80 years (19-96 years) and 51.7% of them were male. There was no significant difference between the study groups in terms of mean age and gender distribution. Endpoint of death was reported in 26.9% of the patients during follow-up; 42 (36.6%) patients in Group A and 13 (14.6%) patients in X Group B died during follow-up period (p= 0.002). The most common etiologies of AKI were sepsis (54.2%, n=109), CCF (14.9%, n=30), medications (13.9%, n=28) and other causes (17.2%, n=34), respectively. Of all patients included, 40.3% had DM, 52.1% had CVD and 22.9% had stage 1, 2 or 3 chronic KD. While there was no significant difference between the two groups with respect to the frequency of DM and CVD, the frequency of CKD was significantly higher in Group A (p=0.01). At hospital admission, Group A had lower mean systolic and diastolic blood pressures and a markedly elevated inotrope requirement (p=0.001). Mean duration of hospital stay was 11.9 ± 8.9 days in the study cohort, being significantly shorter in Group B (mean duration of hospital stay: 12.8 ± 2.2 days in Group A vs. 10.2 ± 2 days in Group B, p<0.001). Univariate analysis of the data obtained from the whole cohort indicated that age ≥65 years, presence of sepsis, history of CVD, inotrope requirement, and the levels of creatinine, CRP, albumin, WBC, NLR and MPV at discharge were all associated with mortality. Multivariate analyses demonstrated that hemoglobin levels at hospital admission and creatinine, CRP, albumin and NLR values at discharge were independent predictors of mortality and the presence of sepsis independently increased the rate of mortality by 5.95 folds [OR: 5.95 (1.25-28.48)). Presence of DM or CVD, the stage of AKI at diagnosis and use of hemodialysis at disease onset or during follow-up were not associated with mortality in neither of the groups nor in the whole study cohort. Baseline NLR values alone were not associated with mortality in both groups, whereas discharge NLR values were significantly associated with mortality in both groups as well as in the whole study cohort (p<0.001). In addition, ΔNLR (discharge NLR – baseline NLR) was significantly associated with mortality in the septic group and in the whole study cohort (p<0.001), but it did not reach the level of statistical significance in the non-septic group. Based on these findings, it may be argued that the change observed in NLR values of the patients during hospitalization is more important than their baseline NLRs and this change is a more valuable predictor of mortality. Conclusion: The present study demonstrated that compared to the baseline NLR values of septic patients, any NLR increase and/or a stable NLR, or even the absence of an insignificant decrease in NLR during follow-up is an important marker of the ongoing inflammatory process in these patients and is apparently associated with the rate of mortality.